PROJECT SUMMARY/ABSTRACT Drug-resistant epilepsy is seriously debilitating and very common, affecting about one-third of the 1-2% of people who experience epilepsy during their lifetime. One of the most effective treatments for drug-resistant epilepsy is dietary therapy, in the form of a very-low-carbohydrate, ketogenic diet. Despite its effectiveness, this diet is not very widely used because of the stringency of the diet and the high commitment required of clinicians and other caregivers. It would be very valuable to understand the mechanism by which altered metabolism produces resistance to epileptic seizures, to ?reverse-engineer? it, and to discover alternative pharmacologic ways of tapping into this potent and apparently unique anti-seizure mechanism. We have identified a mouse model that recapitulates the seizure resistance seen in ketogenic diet, but that involves a mutation in a single gene, Bad. The seizure resistance in this genetic model is due to alteration in brain cell metabolism, with less glucose utilization and better utilization of alternative fuels such as ketone bodies, similar to the metabolic changes on a ketogenic diet. We have also discovered a downstream mechanism that is altered both by Bad alteration and by ketogenic diet: a metabolically-sensitive class of ion channels, the ATP-sensitive potassium channels (KATP channels), become more activated in response to metabolic changes. These channels are critical for seizure resistance of the Bad-altered mice, and we have also found that they are responsible for anti-seizure effects of BAD knockout in a brain slice model of seizure. We have also recently learned that KATP channel activation depends on the expression of the BAD protein in individual neurons, which means that the effects of BAD can be genetically targeted to individual cell types or to specific brain regions. This ability to target the genetic manipulation of the BAD protein ? which cannot be done for a global manipulation like diet ? creates the opportunity to learn the cellular sites of action where BAD modification is required to produce seizure resistance. We now have a conditional knockout allele of the Bad gene (Bad flox/flox) that can be used in combination with various ?driver lines? that express Cre recombinase in specific cells. We will determine whether BAD knockout is effective in slice seizure models or against seizures in mice, when the knockout is restricted to certain targets, for instance, to neurons in specific brain regions like the dentate gyrus that are hypothesized to function as ?seizure gates?. We will also test a pharmacological approach to producing the anti-seizure effects of BAD, by asking whether a specific class of BAD-mimetic compounds is capable of reversing or mimicking the effect of BAD knockout on seizure-like events in slices. These studies will advance our mechanistic understanding of metabolic seizure resistance and more generally of endogenous ?seizure gates?, and will explore new pharmacologic approaches to drug-resistant epilepsy.